Top Banner
AUTOMATION SYSTEM ARCHITECTURE USING OPEN INDUSTRIAL STANDARDS WILLIAM D. FERRAZ, RODRIGO P. PANTONI Smar Equipamentos Industriais Ltda. Av Antonio Furlan Jr. 1622 CEP 14160500 Sertaozinho - SP E-mails: [email protected] , [email protected] Abstract In modern industries the control systems are inserted inside a broader context other then the control problem do- main. This broader context is generally referred to as Automation System, which also addresses peripheral issues like control data management and associated communication protocols, maintenance of the control system components (e.g. sensors and fi- nal control elements) known as ‘Asset Management’, management of data exchange among subsystems, interoperability among devices and among systems, etc. This article shows that most of the previous mentioned issues are addressed using not only elec- tronics and network standards as well as open industrial software standards in the host level (control room workstations) estab- lished upon a common open architecture for automation systems based on SOA (Service Oriented Architecture). Keywords Industrial Standards, SOA, Network of things, Automation System Architecture 1- Introduction In modern industries the control systems are in- serted inside a broader context other then the control problem domain. This broader context is generally referred to as Automation System, which also ad- dresses peripheral issues like control data manage- ment and associated communication protocols, main- tenance of the control system components (e.g. sen- sors and final control elements) known as ‘Asset Management’, management of data exchange among subsystems, interoperability among devices and among systems, etc. This article shows that most of the previous mentioned issues are addressed using not only electronics and network standards as well as open industrial software standards in the host level (control room workstations). Thus, the automation systems evolved in the last decade from stand-alone specialized workstations using vendor-specific (pro- prietary) standards for hardware, software and com- munication protocols to full-fledged hardware and software architecture supporting high-speed data ex- change among all sorts of entities, from biometrics and Internet-aware components to valve positioners (Peluso and Wallace, 2003). This augmentation in the amount of digital-communication enabled devices and the associated data processing generated by the interaction of these devices requires not only the state-of-the-art on software and hardware develop- ment but it requires the knowledge of the state-of-the- art technology to design the system architecture to handle the communication involved in such architec- ture. Dominance over such technologies that can properly deal with the huge amount of information available in the nowadays automation systems net- work is necessary to transform raw information in usable knowledge (Hayes and Alberts, 1996) and also to control the information exchange (Duchastel, 2001) between the network components. In order to guarantee the accomplishment of common factory-floor production and engineering requirements like quality and reliability, stability and safety issues among others, it is necessary to have data and information flowing up and down through the several different layers that compose an enterprise structure, and thus through distinct information- domains e.g. Neural-Network and PID control (elec- tronics engineering domain) affecting production efficiency and planning (Production Management domain). The automation systems community soon realized that it is not effective, and in some cases even not acceptable, to try to siege technology as well as knowledge development and usage by using proprietary solutions or proprietary artifacts. This context conducted the development of non- proprietary solutions. To harmonize different ven- dors' solutions (non-proprietary) for distinct informa- tion domains accomplishing the engineering require- ments, it is common sense that the enterprise automa- tion system might rely upon a common architecture (Emerson, 2005) (Gelle et al, 2003) to enable open standards interoperability. In order to mitigate what have been exposed above, international non-profit organizations play a fundamental role. Some of them, like OPC Founda- tion (OPC – OLE for Process Control), Fieldbus Foundation TM (FF) and IEC (International Electro- technical Commission) cooperate among themselves to enable the development of open and consistent automation system architectures. Examples of such efforts are the FF HSE (High Speed Ethernet) speci- fications, IEC 61804-2 (EDDL - Electronic Device Description Language) and OPC UA (Unified Archi- tecture) specifications (Fieldbus Foundation, 1999a and 199b) (OPC Foundation, 2006). As it's demonstrated in the next paragraphs, an automation system that combine the implementation of such technologies is able to succeed in the hard task of observe, classify and use information to cope with the enterprise ROI (Return Of Investment) and TCO (Total Cost Of Ownership). 63 of 67
5

AUTOMATION SYSTEM ARCHITECTURE USING OPEN INDUSTRIAL STANDARDS

Mar 29, 2023

Download

Documents

Nana Safiana
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Microsoft Word - CLCA_1.docWILLIAM D. FERRAZ, RODRIGO P. PANTONI
Smar Equipamentos Industriais Ltda.
E-mails: [email protected] , [email protected]
Abstract In modern industries the control systems are inserted inside a broader context other then the control problem do-
main. This broader context is generally referred to as Automation System, which also addresses peripheral issues like control
data management and associated communication protocols, maintenance of the control system components (e.g. sensors and fi-
nal control elements) known as ‘Asset Management’, management of data exchange among subsystems, interoperability among
devices and among systems, etc. This article shows that most of the previous mentioned issues are addressed using not only elec-
tronics and network standards as well as open industrial software standards in the host level (control room workstations) estab-
lished upon a common open architecture for automation systems based on SOA (Service Oriented Architecture).
Keywords Industrial Standards, SOA, Network of things, Automation System Architecture
1- Introduction
serted inside a broader context other then the control
problem domain. This broader context is generally
referred to as Automation System, which also ad-
dresses peripheral issues like control data manage-
ment and associated communication protocols, main-
tenance of the control system components (e.g. sen-
sors and final control elements) known as ‘Asset
Management’, management of data exchange among
subsystems, interoperability among devices and
among systems, etc. This article shows that most of
the previous mentioned issues are addressed using
not only electronics and network standards as well as
open industrial software standards in the host level
(control room workstations). Thus, the automation
systems evolved in the last decade from stand-alone
specialized workstations using vendor-specific (pro-
prietary) standards for hardware, software and com-
munication protocols to full-fledged hardware and
software architecture supporting high-speed data ex-
change among all sorts of entities, from biometrics
and Internet-aware components to valve positioners
(Peluso and Wallace, 2003). This augmentation in the
amount of digital-communication enabled devices
and the associated data processing generated by the
interaction of these devices requires not only the
state-of-the-art on software and hardware develop-
ment but it requires the knowledge of the state-of-the-
art technology to design the system architecture to
handle the communication involved in such architec-
ture. Dominance over such technologies that can
properly deal with the huge amount of information
available in the nowadays automation systems net-
work is necessary to transform raw information in
usable knowledge (Hayes and Alberts, 1996) and also
to control the information exchange (Duchastel,
2001) between the network components.
In order to guarantee the accomplishment of
common factory-floor production and engineering
requirements like quality and reliability, stability and
safety issues among others, it is necessary to have
data and information flowing up and down through
the several different layers that compose an enterprise
structure, and thus through distinct information-
domains e.g. Neural-Network and PID control (elec-
tronics engineering domain) affecting production
efficiency and planning (Production Management
domain). The automation systems community soon
realized that it is not effective, and in some cases
even not acceptable, to try to siege technology as
well as knowledge development and usage by using
proprietary solutions or proprietary artifacts. This
context conducted the development of non-
proprietary solutions. To harmonize different ven-
dors' solutions (non-proprietary) for distinct informa-
tion domains accomplishing the engineering require-
ments, it is common sense that the enterprise automa-
tion system might rely upon a common architecture
(Emerson, 2005) (Gelle et al, 2003) to enable open
standards interoperability.
above, international non-profit organizations play a
fundamental role. Some of them, like OPC Founda-
tion (OPC – OLE for Process Control), Fieldbus
Foundation TM
to enable the development of open and consistent
automation system architectures. Examples of such
efforts are the FF HSE (High Speed Ethernet) speci-
fications, IEC 61804-2 (EDDL - Electronic Device
Description Language) and OPC UA (Unified Archi-
tecture) specifications (Fieldbus Foundation, 1999a
and 199b) (OPC Foundation, 2006).
As it's demonstrated in the next paragraphs, an
automation system that combine the implementation
of such technologies is able to succeed in the hard
task of observe, classify and use information to cope
with the enterprise ROI (Return Of Investment) and
TCO (Total Cost Of Ownership).
63 of 67
Implement a Common Network
structure is the basic source of the Automation Sys-
tem's raw data. In the modern plants, however, the
data generated varies from simple measurements like
a pressure value to complex diagnostics information
like a valve signature 1 or even measurement subsys-
tems. Analog technologies are not suitable to transmit
this amount of data, so the technology evolved to-
wards digital-processor field devices in the last two
decades.
disappear all the time. Often, those who buy into the
'technology de jour' 2 have later been disappointed
and incurred in unnecessary expense to replace a
'promising' but unsupported technology.
tion technology architectures, users state their con-
cerns a hundred different ways, but in the end, what
users (Business and Process) seek are assurances the
technology platform they choose provides (Zielinski,
2004):
tation and equipment independent of the
Host - valves, transmitters, motor starters,
remote IO, etc.;
tion and equipment is engineered;
• Flexibility and efficiency in how plant-floor
data are shared throughout the enterprise;
• Ease of maintenance;
ers have long-term commitments to expand
and improve the underlying technology;
• Easy access to production data;
• Easy connectivity throughout the business
endpoint;
Among several solutions, the non-profit organi-
zation Fieldbus FOUNDATION TM
enterprise's infrastructure network mentioned above.
As shown in the figure 1, the FF HSE system archi-
tecture provides a framework for describing the
automation system as a collection of physical devices
interconnected by a fieldbus network (Fieldbus
Foundation, 1999a).
sistent hardware and software architecture envisaging
an optimum integration with the end-user, through a
layer called ‘User-Layer’, which, in the extent given
1 Valve signature plots actuator pressure versus travel
(actuator stem position variation over time) 2 French expression. Technology-of-the-day, not proven-
in-use technology
such as MODBUS.
The FF systems' architecture is composed of
two parts. The (low-speed) factory-floor part, named
‘H1’ and the systems' backbone communication net-
work named ‘HSE’, taking the acronym from the
Ethernet's definition High Speed Ethernet, since the
communication relies upon the well known and ac-
cepted Ethernet standard.
FF HSE User layer uses the Electronic De-
vice Description Language (EDDL) to describe the
devices throughout the FF-HSE-based system. EDDL
is a text-based language for describing the digital
communication characteristics of intelligent field
instrumentation and equipment parameters—device
status, diagnostic data, and configuration details—in
an operating system and Human Machine Interface
(HMI) neutral environment.
ting device software drivers to work after a host plat-
form OS upgrade. EDDL was specifically designed
for the EDD (Electronic Device Description) file to
avoid this issue. The IEC 61804-2 standard is not
only OS (Operational System) and HMI independent,
but also fieldbus communication protocol neutral.
Today EDDL technologies establishes the
engineering and operating base on which all major
digital fieldbus protocols—FOUNDATION Field-
bus™, HART®, and PROFIBUS—construct para-
metric and device descriptions. And, because EDDL
is an open technology with international standard
status, it can be easily and effectively applied to any
device and any fieldbus protocol (Emerson , 2005).
From an end-user perspective, it's equally
reassuring to know that a host workstation change
from any OS to any other without require modifica-
tion of the existing EDD files. It is possible to men-
tion two typical examples like: migration of the host
platform (OS) from Windows NT to Windows 2003
Server or from Windows to Linux.
Allying the FF HSE architecture and EDDL
capability is possible to integrate not only the com-
munication systems between networks but it's possi-
ble to integrate the device manufacturer know-how to
the automation infrastructure. However, the automa-
64 of 67
tion system vendor is not limited to the device manu-
facturer know-how. It can also join its own knowl-
edge to it.
tive project to extend the capabilities of the IEC
618342 standard.
add robust organization and graphical visualization
of device data as well as support for persistent data
storage (i.e., permanent data storage) features to IEC
61804-2. Sophisticated devices. As it is protocol and
technology independent (different from other tech-
nologies with the same purpose like FDT - Field De-
vice Tool (Merrit, 2002)), it is even more efficient
when using FF-HSE, with its user-layer, to empower
data and information transfer.
4- OPC AND OPC-UA
HMI software manufacturers have to develop only a
driver for communication with devices, differently of
the way before, which each manufacturer had to de-
velop proprietary drivers for supporting its devices.
A common analogy for OPC DA (Data Ac-
cess) is printer drivers in DOS versus Windows. In
DOS, the application developer had to write a printer
driver for every printer. So AutoCAD, WordPerfect,
and Netscape all had to write a separate printer driver
for every printer they wanted to support. In industrial
automation, companies like Rockwell Automation,
ICONICS and Citect wrote their own HMI software
and a proprietary driver for each industrial device
including every PLC (Programmable Logic Control-
ler) brand in order to retrieve process data.
The standardization of these drivers was de-
veloped by a group of manufactures that have im-
proved the OPC model due to them additions based
on past experiences. It is mentioned for example the
navigation system of the OPC tags through tree struc-
ture, and its division by sectors.
Since 2003, OPC Foundation and a group of
IHM and device manufactures have been working in
a joint venture to define and implement a new OPC
Specification. This new Specification is the OPC UA.
The purpose of OPC UA is to provide en-
hancements for existing and next generation OPC
products in the areas of security, reliability, and in-
teroperability. OPC UA is designed to unify existing
OPC specifications such as DA, DX (Data Ex-
change), HAD (Historical Data Access), and XML
(eXtensible Markup Language) DA into an environ-
ment that will leverage Web-based technologies and
standards such as Web Services, WSDL (Web Ser-
vices Description Language), XML and SOAP (Sim-
ple Object Access Protocol).
Oriented Architecture) through Web Service that
transports XML data, which provides the communi-
cation among different software of different plat-
forms, providing interoperability (state of the art in
IT) (OPC Foundation, 2006).
change in the Internet that the World Wide Web
Consortium (W3C) developed in 1998 (Roy and Ra-
manujan, 2000). XML lets interoperability of differ-
ent software of different platforms to transfer infor-
mation using a common language. Since XML is a
meta-language, it is possible to create new languages
to make a standard talk. In Mathematics for example,
a XML based language called MathML is in use in
the same way OPC uses SOAP (Gelle et al, 2003).
The OPC Foundation has joined the interna-
tional cooperative team of the three leading fieldbus
organizations, the Fieldbus Foundation TM
(FF),
PROFIBUS NutzerOrganisation e.V. (PNO), to ex-
tend the reach of electronic device descriptions
(EDDs) into the OPC unified architecture. This work
provides the total integration between OPC UA and
EDDL to describe data to the host reads, i.e., OPC
Clients have access to complex device data with
automatic integration. This lets an easier and cheaper
development, which brings a better quality for the
product to the final user (Fieldbus Foundation, 2006).
Thus, OPC UA is the key to improve the
transformation of simple data in knowledge, due to
the open high-level mechanism, open the doors to
store the information by an easer way for MES
(Manufacturing Execution System) and ERP (Enter-
prise Resource Planning) applications.
XML
operable automation system in order to establish a
common architecture, other features are necessary.
These other requirements are necessary because FF-
HSE, EDDL and OPC do not provide resources, for
instance, to implement binary and sequential logic as
well as enterprise representation (which are better
provided by the IEC 61131 standard and ISA S-88).
Then it is very clear that another technology
is necessary to integrate all the different standards
used in the automation system. As can be inferred
from the previous sections, the best approach to pro-
mote integration of data, communication and even
graphical elements is the usage of XML (W3C, 2006)
(and its variations like XML Schema, XSLT - eXten-
sible Stylesheet Language Transformations, etc).
XML is used to integrate different software layers as
well as the software in the same network layer be-
cause it is also open (in fact it is public), platform
independent and has a great number of programs and
APIs that allow easy manipulation (read and write
operations) of XML files. Even other specifications
are being produced having XML as basic element of
65 of 67
taleán, 2003) (Gelle et al, 2003).
Thus, XML can directly consolidate not
only device internal data (described, for instance, by
EDDL, CFF - Capability File Format and CFH - Ca-
pability File for FOUNDATION Fieldbus TM
HSE)
and visualization tools (IEC61131, ISA-S88, SGML,
SVG). Having XML as the common language being
used by the different components of the system, it is
possible to extend the cognition to the upper layers of
the automation system facilitating the integration of
the system with business endpoints, MES, etc. The
figure 2 represents the SOAP / XML as the integra-
tion technology.
Figure 2: SOAP-XML as the core ‘language’ of the Common
Automation Architecture
Architecture, can assure long-term integration of the
Manufacturing Facility with other business endpoints,
since SOAP-XML is currently being applied in all
sorts of business automation structures.
6- CONCLUSIONS
new challenge to the mankind, which is the ability to
construct networks of virtually anything. Some has
called this as ‘network of things’ (Sang, 2006). This
new reality, as shown in figure 3, will need not only
new network approaches but also solutions to deal
with such amount of information exchanged through-
out the network. The new ‘network of things’ sce-
nario will be constructed on devices that have an en-
hanced cognition ability, using SOAP /Web Services,
characterizing the ‘Total Information Age’.
The first implementations of systems that
are using these technologies have demonstrated sev-
eral benefits not only for technicians but also for en-
terprise managers and business agents.
Figure 3: The ‘Network of Things’
According to ARC (The ARC Strategies
Report, 2001), one set of the benefits is shown in the
following graph (Figure 4), which guarantee better
ROI figures than conventional, proprietary systems
(Verhappen, 2005).
Source: ARC
to choose not only the best control strategy (or algo-
rithms) but also the components and associated soft-
ware used to implement the control system itself and
its insertion in the enterprise automation. This means
that the gains obtained by using a better control sys-
tem algorithm will not be evident if a sensor used in
the implementation is affected by defects not quickly
detected. Using modern software the resolution of
such defects could be done faster than using legacy
software and equipment, reducing the downtime and
consequently, reducing unplanned breakdown.
scenario this article has shown that the Automation
System of the Total Information Age should be able
to map the intricacies of its problem domain into
SOAP/XML using open standards, constructing a
common open architecture. It was shown that EDDL,
FF-HSE and OPC UA are able to accomplish such
task. TCO is then mitigated since this approach will
prepare the enterprise to face the nowadays needs and
also the future fractal network structure, where the
users will have almost no concern about the network
and software technology being used but they will be
concerned about the services being offered.
66 of 67
HSE High Speed Ethernet
DCS Distributed Control System
DOM Document Object Model
EDD Electronic Device Description
ERP Enterprise Resource Planning
FDT Field Device Tool
FF Fieldbus FOUNDATION TM
HCF HART Communication Foundation
HMI Human Machine Interface
HSE High Speed Ethernet
IEC International Electrotechnical Commission
ica)
OPC DA Data Access
OPC DX Data Exchange
OPC UA Unified Architecture
SOA Service Oriented Architecture
SVG Scalable Vector Graphics
XML eXtensible Markup Language
mations
REFERENCES
mation and Control Systems.
Digit.hvm. ISSN 1575-2275.
Fieldbus Foundation (1999a). FF-581-1.3: System
Architecture.
Overview.
Industrial Automation.
Information Dominance: Beyond Information War.
Merrit, R (2002). FDT Is Not a Fieldbus Panacea.
Control.
gence: Hydrocarbon Engineering.
language. IT PRO. p. 32-36.
Sang S. (2006). Webservices course at
<www.javapassion.com/webservices> .
guese). - Monograph presented to UniCOC – Ribei-
rao Preto.
sertation-USP.
Sensor Strategies for the E-World.
Verhappen I. (2005). Fieldbus – Status and Future
Directions. FF EUC, Buenos Aires.
W3C (2006). Extensible Markup Language.
<www.W3.org/XML>.
ity.